Why applications should drive your IT investments

By | 4 minute read | January 20, 2021

As business and IT leaders strive to optimize resources and reduce costs, it can be tempting to choose a single standard server to deploy for their entire organization. Commodity servers are perceived as low-cost and effective infrastructure, so it seems like a smart investment.

However, this temptation can yield the opposite results as intended.

Not only can it be expensive and time-consuming to migrate applications to a single standard server, but doing so can inadvertently introduce new issues, as different applications can have different infrastructure requirements. This is why it’s important to analyze an application’s total cost of ownership (TCO) before making this type of investment.

When all application requirements and data center costs are considered, commodity scale-out servers can actually drive up the total cost of IT spend and fail to deliver ROI. Meanwhile, after performing this analysis, the IBM Z® platform emerges as a superior infrastructure option for many applications.

Let’s look at offload assessments performed by IBM IT Economics that help you see how IBM Z can lower your IT infrastructure costs.

How the evolution of application lifecycle management impacts TCO

IT applications – and the business processes implemented within them – represent the real assets of modern organizations. This is why organizations still run applications that were originally deployed more than half a century ago.

But application lifecycle management (ALM) has changed over the years, and organizations have adopted new trends while maintaining already-deployed applications. This has created different layers of technologies, and application integration has become a crucial challenge for every organization.

Application development and maintenance requires significant resources. And, after a few years, the total cost of an application becomes much larger than the cost of the infrastructure it runs on.

How IBM Z can lower TCO and improve ROI

Many companies find that continuing to run their applications on IBM Z yields a lower TCO and higher ROI than new ALM techniques that promise a lower TCO for applications running on other systems.

Here’s an example: At the request of clients, IBM IT Economics performed TCO offload assessments involving rewrites of up to 10 million lines of application code. The assessment results included:

  • Clients had—on average—a 2x lower annual TCO keeping their applications on IBM Z versus moving to an x86-based infrastructure[1]. This was mainly due to underestimating application migration costs, parallel environment maintenance period costs and the sizing of the equivalent x86 infrastructure once fully deployed.
  • Offload projects ran beyond their planned completion date and budget as well as falling short of the project’s planned scope due to the complexity of migrating applications. Even in cases considered technically successful, analysis found that the project either faced a long ROI break-even point of 20 or more years, or none whatsoever.

The risks of long project duration and high cost tend to be why many companies avoid application migration and ultimately keep their applications running on IBM Z. But many also find cost savings through optimization of their IBM Z environment or by exploring new service provisioning models such as shared data centers or IBM Z cloud offerings.

Several characteristics of IBM z15™ enable solutions to meet the most challenging business requirements while minimizing IT costs.

These advancements, combined with investments made in IBM Z applications running in thousands of enterprises worldwide, have convinced many IBM Z clients to further enhance and expand their mainframe environments. Instead of replatforming their applications to another hardware architecture, you can applications and modules on IBM Z that extend the capabilities of their original applications.

Below are some additional examples where IBM Z provides a cost-effective TCO case.

Java® applications through the exploitation of IBM Z specialty engines

  • Workloads can leverage IBM Z Integrated Information Processors (zIIPs) and the integration of JVM with IBM z/OS® to minimize general processor compute charges and software license charges by offloading the work to zIIPs.
  • Workloads can also leverage Integrated Facility for Linux® (IFLs) on IBM Z. Because Java workloads can be densely consolidated onto IBM Z, requiring fewer processor cores than on x86 servers, middleware software costs for both the z/OS and Linux on IBM Z environments can be significantly reduced.
  • Workload consolidation analysis from 17 IBM IT Economics assessments found that the same Java workloads on IBM LinuxONE™ or IBM Z provided—on average—a 54% lower TCO over five years than on compared x86 servers[2].

Competitive database solutions

  • The number of software licenses can be reduced by as much as 78%.
  • Sizing analysis from IBM IT Economics assessments of clients with business-critical loads show most x86 Linux workloads have a core consolidation ratio ranging from 10 to 32.5 distributed cores to one IFL with an average of 17x fewer cores[3].

Data warehouse and transactional business intelligence solutions:

  • If the master copy of the data already resides on the IBM Z system in IBM z/OS and in Linux on Z partitions for both structured data and big data repositories, these applications are already co-located with the data, eliminating the need to support off-platform environments.
  • While TCO differences vary for these solutions depending on each of the client’s requirements, workload footprints are centralized on a single platform, bringing infrastructure savings and efficiencies with reduced latency.

The Benefits of IBM Z

IBM Z enables legacy and open environments to coexist on the same hardware platform so that businesses can streamline operations and optimize costs. IBM Z can provide a lower TCO compared to alternative scale-out solutions while enabling applications to exploit the latest development and delivery approaches on an enterprise proven infrastructure.

>> Discover IBM z15

[1] IBM customers across different industries and geographies requested TCO analysis of their IBM Z workload offload projects. Mainframe operations ranged in size from 88 to 12,500 MIPS and required some application rewrite effort, varying from 750,000 to 10,000,00 lines of code, to move to an x86 environment. Client workloads were comprised of IBM monthly license charges (MLC) and International Program License Agreement (IPLA) licensing and independent software vendor (ISV) licensing. Hardware was comprised of IBM Z servers running z/OS and specialty engines such as IBM z Integrated Information Processors (zIIPs). Each client engaged the IT Economics team to evaluate the workloads, their existing mainframe environment and proposed distributed environment for the offload. One third of clients had already initiated IT offload activities while another third had completed the effort, although reported the project as a failure. The remaining third was considering offload and was still in the planning phase of their project. For all the TCO assessments, IT Economics consultants met on-site with the client to discuss offload planning and execution, analyzed forecasted project costs, and examined actual cost to date for those in execution mode. IT Economics analysis observed activity omissions and underestimated sizings in the offload projects and quantified offload costs for the clients. The clients concurred that their plans had underestimated the effort, cost and risk of their offload project plans. The average five-year x86 TCO for all clients was 3.2x higher than the IBM Z TCO, with a range of 2.1x to 3.7x. [2] 17 IT Economics assessments involving analysis of Java x86 workloads for consolidation onto zIIPs or IFLs on IBM Z, or LinuxONE were selected from diverse industries (35% financial, 25% government, 12% healthcare, 6% retail, 6% telecommunications, 6% utilities) and geographies. The assessments included were performed for clients with business-critical workloads running in production and non-production environments. The workloads targeted for consolidation from x86 and distributed servers were IBM Java application server middleware running on different types of x86 and distributed servers. TCO costs included hardware, software, networking, energy, floor space and people costs. TCO savings with zIIPs, IFLs or LinuxONE ranged from 20% to 85% over five years with an on average savings of 54%. Each client engaged the IT Economics team to evaluate the distributed workloads and the proposed IFL or LinuxONE environment for the consolidation. For each assessment, IT Economics consultants met with the client to discuss consolidation planning and execution, analyzed the client’s current total cost of ownership, and provided a projected total cost of ownership with workload consolidation based on estimated core consolidation ratios for the client’s workloads. For additional information on x86 workload analysis contact the IBM IT Economics team, it.economics@us.ibm.com. [3] 20 IT Economics assessments involving analysis of x86 workloads for consolidation onto IFLs on IBM Z or LinuxONE were selected from diverse industries (35% financial, 25% government, 5% healthcare, 10% retail, 10% technology, 10% transport, 5% utilities), and different geographies (5% North America, 15% Latin America, 30% Europe, 20% Asia Pacific, and 30% Greater China Group). The assessments included were performed for clients with business critical workloads running in production and non-production environments. The workloads targeted for consolidation from x86 and distributed servers were IBM and third party proprietary and open source databases, application server middleware and industry specific solutions running on different types of x86 and distributed servers. Each client engaged the IT Economics team to evaluate the distributed workloads and the proposed IFL or LinuxONE environment for the consolidation. For each assessment, IT Economics consultants met with the client to discuss consolidation planning and execution, analyzed the client’s current total cost of ownership, and provided a projected total cost of ownership with workload consolidation based on estimated core consolidation ratios for the client’s workloads.